The present invention relates to a method of backing up data stored in a storage unit provided in a computer system. More particularly, the present invention relates to a method of backing up data from a disk system into a storage medium for a backup without having to interrupt I/O processing of an application program running on a computer.
A typical computer system needs to back up data stored in a storage medium such as a disk system for protecting the data from being broken by some unfavorable factors such as system failure, human inadvertent mistake, and disasters. In general, an ordinary user backs up the data at night once a day or a week by interrupting an online service. In recent days, however, a request for an uninterruptible online service is rising for the purpose of globalizing an enterprise or enhancing a customer""s satisfaction level. For this purpose, a time to be spared for a data backup is made shorter and shorter.
As a technology for coping with such a request, there is known a technology disclosed in the white paper xe2x80x9cLegato NetWorker for EMC Symmetric Installation and Administration Guidexe2x80x9d of the Legato, Ltd. disclosed in a website xe2x80x9cwww.legato.comxe2x80x9d on the Internet. The technology disclosed therein is arranged to provide a pair of disk drives in a disk system. In the ordinary online service, the data is duplicated and stored in these two disk drives. When the data stored in one disk drive (corresponding to a user volume) is updated by the online service, the updated content is reflected on the other disk drive (corresponding to a secondary volume). In backing up the data, in response to an instruction given from a host computer, the system is operated to write the data that is not reflected on the disk drive and held in a cache memory (the data is called dirty data), onto a disk drive, split a pair of disk drives into each one, and avoid reflection of the updated data in the user volume onto the secondary volume. Then, the host computer for performing the online service operates to continue the processing with the user volume. The secondary volume stores the data at the time of splitting the pair of disk drives. This data is copied onto a backup storage unit such as a magnetic tape device by means of another computer, for the purpose of backing up the data. Upon completion of the backup processing, the host computer operates to issue a re-synchronizing command so that it may return the user volume and the secondary volume to the duplex state.
This backup method makes it possible to obtain a consistent backup of the data because all the data is frozen at a time when all the dirty data has been reflected onto the primary volume. Further, the split of the pair state is completed for quite a short time, so that no interruption of the online processing is required. The data (frozen data) held in the secondary volume at the time of the pair split is called snapshot.
As another method of obtaining the snapshot, there are known a technology disclosed in the white paper xe2x80x9cCelestra Architecture for Serverless Backupsxe2x80x9d and xe2x80x9cCelestra Block Copy Interface Specificationxe2x80x9d of the Legato, Ltd. disclosed in the website xe2x80x9cwww.iguard.comxe2x80x9d on the Internet. The technologies disclosed therein utilize a command for copying data from one volume to another volume or from a volume to a data device. This command is called a CCOPY command. In advance of obtaining a backup, the host computer operates to flush the dirty data and then write the data into a user volume. Then, the host computer uses the CCOPY command to copy the data from the user volume into the work volume. When the application program running on the host computer operates to write the data onto an area where no data is copied of the working volume, the copy program operates to detect it and keep the writing process waiting until the copy of the area is completed. The foregoing processing makes it possible to obtain the data at the time of starting the process of obtaining a backup process, on the working volume. After that, the host computer operates to use the CCOPY command for backing up the data from the working volume into the tape device.
In order to prevent the application program from being hung up as a result of a timeout or reduce the waiting time of the online processing, it is necessary to make the waiting time of the application program minimum. Through the use of a method of obtaining the snapshot according to the foregoing prior art, when the application program tries to update the data on the area that is not copied into the working volume, the writing processing is kept waiting until the copy on the area is completed. In copying all the data of the user volume onto the working volume at a batch, the processing of the application program is kept waiting until copying of all the data is completed. In order to avoid such a shortcoming, it is considered that the target data is copied into the working volume bit by bit. However, in the case of copying the data bit by bit, one CCOPY command is used for partially copying the area to be backed up. It means that the number of the issuing times of the copy command is increased. Further, the foregoing prior art is arranged to temporarily copy the data of the user volume into the working volume and then save the data into the storage medium for a backup such as a magnetic tape. Hence, the time up to the completion of the backup processing is made longer.
It is an object of the present invention to overcome those shortcomings of the prior arts involved and to provide a method of backing up the data which is arranged to reduce the waiting time of the writing process of the application program as reducing the times of copies.
In achieving the object, according to an aspect of the invention, a user data used by a computer in a storage unit system is held dually in first and the second storage units. The storage system operates to update the user data of the first storage unit and to reflect the updated content onto the second storage unit in response to a request for updating the user data given from the computer. In backing up the data, a splitting instruction for releasing the duplex state of the first and the second storage units is given from the computer into the storage system. In response to the splitting instruction, the storage system operates to interrupt reflection of the updated user data in the first storage unit onto the second storage unit. Then, a copy instruction for copying the user data held in the second storage unit into the backup storage device is sent from the computer into the storage system. In response to a copy instruction, the storage system operates to transfer the user data held in the second storage unit into the backup storage device.
According to another aspect of the invention, in backing up the data, a copy request for a specified area to be copied is issued from the computer into the storage system. In response to the copy request, the storage system operates to release the duplex state of the first and the second storage units and start copying of the user data from the second storage unit into the backup storage device. Preferably, when the duplex state is released, a report on completion is given back from the storage system into the computer. The computer operates to restart the update processing of the user data in response to this report.
According to another aspect of the invention, the storage system provides the user volume for storing the user data to be used by the computer as well as a working volume for saving the data in copying the data of the user volume into the backup storage device and backing up the data. In response to the copy request for the data to be copied from the user volume into the backup storage device, the storage system operates to read the user data held in the area specified by the copy request out of the user volume and then to start the copy process to the backup storage device. During the copying of the user data onto the backup storage device, when a request is given for writing the data onto the user data in the area specified by the copy request, the user data held in the area where the data is written is transferred onto the working volume according to the writing request. After the user data is saved in the working volume, the data writing is done upon the writing request. The storage system operates to determine if the user data to be copied into the backup storage device is held in an area inside of the user volume or an area inside of the working volume, read the user data based on the determined result, and transfer the user data into the backup storage device.
A further understanding of the nature and advantages of the invention herein may be realized by reference to the remaining portions of the specification and the accompanying drawings.